In the home stretch to next week’s election, a number of articles have attempted to rebut claims that the Trump administration has practiced “phony” populism. But the only piece of real-world evidence these articles cite to defend the Trump administration’s record in helping working-class voters turns out to be either false or highly misleading.

Specifically, one of the articles defending the Trump record, by Alan Tonelson, highlights wage growth in “pivot counties”—counties that voted for Obama twice but then voted for Trump—and claims that “Average annual private-sector pay in most of these [pivot] counties rose faster during the first three years of the Trump administration than during the last three years of the [sic] Mr. Obama’s presidency.”

In Tonelson’s telling, this wage growth justifies the vote-flipping in those counties between Obama and Trump because the Trump administration has done something that has boosted wage growth in these presumably blue-collar counties. But Tonelson’s analysis is wrong, for a number of reasons.

First, our calculations show that pivot counties didn’t see faster wage growth on average. As Figure A shows, between 2013 and 2016 average real annual pay in pivot counties grew by 4.3%, and between 2016 and 2019 these pivot county average earnings grew by just 2.2%. In all other (nonpivot) counties, the slowdown in earnings growth was smaller: Average earnings grew by 4.0% in the first period and then 3.1% in the second period.

It is true, as Tonelson notes, that a majority of pivot counties had faster nominal wage growth in the Trump period: By our calculations, 57% of pivot counties had faster nominal growth in the Trump period. But pivot counties actually fared worse than other counties: 67% of nonpivot counties also saw faster nominal growth. In addition, after accounting for inflation, only 33% of pivot counties saw faster real average pay rise. Nonpivot counties fared slightly better: 43% of nonpivot counties had faster real pay growth.

In short, average wage growth in pivot counties fell further behind nonpivot counties during the Trump administration, regardless of whether or not one accounts for inflation. In fact, inflation-adjusted wage growth for pivot counties in the second Obama administration was faster than for nonpivot counties, yet this pattern outright reversed during the Trump administration.

The lack of evidence supporting claims that the working class has prospered under the Trump administration should shock nobody—the policy record is clear that the administration’s claims to care about working-class voters’ economic plight were indeed phony. The Trump administration prioritized tax cuts for corporations, repealing access to affordable health coverage, and wrecking any institution or labor standard that gives typical workers leverage or bargaining power. The administration also did not push to raise the minimum wage, but did push to appoint hard-money cranks and reliable conservative partisans to the Federal Reserve.

It is true that some pockets of strength in wage growth have appeared during the Trump administration. But these can mostly be explained by state-level actions in raising minimum wages and a continued tightening of the labor market before 2020—a steady improvement that Trump inherited from the previous administration and which can be chalked up mostly to the Fed’s admirable forbearance in raising interest rates.

It is certainly true that working class voters’ economic interests were disproportionately ignored (or even constrained) before the Trump administration, but this disadvantage only increased during the past four years. Given the policy record of the Trump administration, this should come as no surprise—everything they have tried to do has privileged the already-rich and harmed everybody else.